Interim 4′-[methyl-¹¹C]-thiothymidine PET for predicting the chemoradiotherapeutic response in head and neck squamous cell carcinoma: comparison with [¹⁸F]FDG PET

We conducted a retrospective analysis of prospectively collected data. The prospective study consisted of 259 consecutive, untreated patients with primary head and neck tumors who underwent [¹¹C]4DST PET/CT study between May 2011 and March 2020. All patients included gave written and informed consent and study protocol was approved by the institutional ethics committee of the Kagawa University (registration number: 23011). From these patients, 32 patients (29 males, 3 females; mean age, 63.3 years; age range, 51–74 years) were selected for this retrospective analysis. Eligible patients fulfilled all of the following criteria: (1) histologically confirmed squamous cell carcinoma of the head and neck region with treatment-naïve; (2) both [¹¹C]4DST and [¹⁸F]FDG PET/CT performed before treatment (baseline); (3) received radical chemoradiotherapy after baseline PET/CT studies; and (4) both [¹¹C]4DST and [¹⁸F]FDG PET/CT performed at approximately 40 Gy point during chemoradiotherapy (interim). This retrospective data collection was compliant with the institutional ethics committee of the Kagawa University, with a waiver of informed consent (registration number: 2020078). Some of the data from 18 of these patients were used in a previous study [9]. Their clinical data are summarized in Table 1.

Radiotherapy was administered to the primary head and neck regions once daily using 4–MV photons with a pair of bilaterally opposed fields in the upper neck and an anterior port at the lower neck. Patients were irradiated with a total dose of 62–70 Gy in 2 Gy fractions once daily. After administration of 40 Gy, the clinical target volume was reduced to encompass only the primary tumor and involved neck lymph nodes. All patients received 1–3 courses of systemic chemotherapy. Ten patients received chemotherapy with cisplatin (70 mg/m²) and 5-fluorouracil (1,000 mg/m² continuous infusion for 5 days), and 20 patients received chemotherapy with nedaplatin (80 mg/m²) and S-1 (100 mg/day for 14 days). The remaining 2 patients with T2 laryngeal cancer received weekly docetaxel (10 mg/m²) chemotherapy 6 times during radiotherapy.

Response at 3 months after completion of chemoradiotherapy was clinically evaluated on the basis of naso-endoscopy (25 patients), CT (26 patients), magnetic resonance imaging (1 patient), [¹¹C]4DST and [¹⁸F]FDG PET/CT (32 patients), and biopsy (2 patients). Patients without or with evidence of residual or recurrent disease at 3 months were classified as showing a complete response (CR) and non-CR, respectively, by the head and neck tumor board team meetings, prospectively.

The radiotracers, [¹¹C]4DST and [¹⁸F]FDG, were manufactured using an automated synthesis system with HM-18 cyclotron (QUPID; Sumitomo Heavy Industries Ltd, Tokyo, Japan). The [¹¹C]4DST was synthesized using the method described by Toyohara et al. [6].

All acquisitions were performed using a Biograph mCT 64-slice PET/CT scanner (Siemens Medical Solutions USA Inc., Knoxville, TN, USA), which has an axial field of view of 21.6 cm. Interim PET/CT scans were obtained at approximately 40 Gy point during chemoradiotherapy (median 42 Gy; range 32–50 Gy). The median intervals between [¹¹C]4DST and [¹⁸F]FDG PET/CT studies for baseline and interim scans were 5 days (range 0–70 days) and 1 day (range 0–44 days), respectively.

Patients fasted for at least 5 h prior to [¹⁸F]FDG administration, and a normal glucose level in the peripheral blood was confirmed prior to [¹⁸F]FDG injection. Emission data were acquired from the midcranium to the proximal thighs (2 min per bed position) at 15 min after intravenous injection of [¹¹C]4DST (7.4 MBq/kg) and 90 min after intravenous injection of [¹⁸F]FDG (3.7 MBq/kg). Unenhanced, low-dose CT of the same area was performed for attenuation correction and image fusion. PET data were reconstructed with an ordered subset expectation maximization algorithm, incorporating correction with point-spread function (PSF) and time-of-flight model (2 iterations, 21 subsets) using a Gaussian filter. Quantification were based on PSF-reconstructed data.

A board-certified nuclear medicine physician, who had 7 years of experience in reading [¹¹C]4DST and [¹⁸F]FDG PET/CT, performed PET/CT image analyses retrospectively. PET/CT image were assessed on the presence of foci of increased activity within the primary tumor greater than surrounding background. The standardized uptake value (SUV) was calculated using the following formula: SUV = c_dc/(d_i/w), where c_dc is the decay-corrected tracer tissue concentration (Bq/g); d_i, the injected dose (Bq); and w, the patient’s body weight (g). The maximum SUV (SUVmax) from both the [¹⁸F]FDG and [¹¹C]4DST PET studies and metabolic tumor volume (MTV) from [¹⁸F]FDG PET or proliferative tumor volume (PTV) from [¹¹C]4DST PET for primary tumor were measured. MTV or PTV was defined as the volume with an SUVmax greater than 2.5 and to exclude adjacent [¹⁸F]FDG-avid or [¹¹C]4DST-avid structures [8, 9]. When no tumor-related radioactivity was discernible visually (interim PET studies), the mean SUV of the primary region on the basis of baseline PET studies was measured and MTV or PTV was assumed to be zero. Total lesion glycolysis (TLG) and total lesion proliferation (TLP) were calculated in the [¹⁸F]FDG and [¹¹C]4DST PET studies, respectively, as follows: MTV or PTV × mean SUV. The differences in SUVmax (ΔSUVmax), MTV (ΔMTV) or PTV (ΔPTV) and TLG (ΔTLG) or TLP (ΔTLP) from baseline to interim PET scans were calculated using the following formula: ΔSUVmax = (interim SUVmax − baseline SUVmax) × 100/baseline SUVmax; ΔMTV = (interim MTV − baseline MTV) × 100/baseline MTV; ΔPTV = (interim PTV − baseline PTV) × 100/baseline PTV; ΔTLG = (interim TLG − baseline TLG) × 100/ baseline TLG; ΔTLP = (interim TLP − baseline TLP) × 100/ baseline TLP.

The data were analyzed using SPSS statistical software (version 26; IBM). PET parameters of the baseline and interim scans were compared using the paired t test. PET parameters between CR and non-CR groups were compared using the Mann–Whitney U test. Receiver operating characteristics (ROC) analysis was performed to determine the effectiveness of PET parameters for differentiating the early chemoradiotherapeutic response. Two-tailed values of P < 0.05 were considered statistically significant.

Primary tumors were detected in all patients on both the [¹¹C]4DST and [¹⁸F]FDG baseline PET images. All patients showed increased uptake on [¹⁸F]FDG interim PET images, whereas 18 showed no increased uptake in the primary region on [¹¹C]4DST interim PET images. The results in baseline and interim PET parameters are presented in Table 2 and Fig. 1. [¹¹C]4DST SUVmax, PTV, and TLP for interim were significantly lower than the corresponding values for baseline (all P < 0.001) (Table 2). [¹⁸F]FDG SUVmax, MTV, and TLG for interim were also significantly lower than the corresponding values for baseline (P < 0.001, = 0.02, and = 0.005, respectively).

After chemoradiotherapy, 25 patients were found to show CR and 7 non-CR. Table 3 summarizes the results of the association between PET parameters and therapy response. [¹¹C]4DST ΔSUVmax, ΔPTV, and ΔTLP for CR group showed significantly greater reduction than the corresponding values for non-CR group (P = 0.044, < 0.001, and < 0.001, respectively), whereas there were no significant differences in [¹⁸F]FDG ΔSUVmax, ΔMTV, or ΔTLG between CR group and non-CR group.

[¹¹C]4DST ΔSUVmax, ΔPTV, and ΔTLP demonstrated good performance for differentiating the early chemoradiotherapeutic response, with AUC values of 0.75, 0.91, and 0.89, respectively (P = 0.044, < 0.001, and = 0.002, respectively) (Table 4 and Fig. 2). [¹¹C]4DST ΔPTV of -90 was the best cutoff value for early identification of patients with non-CR. [¹⁸F]FDG ΔSUVmax, ΔMTV, and ΔTLG demonstrated poor performance for differentiating the early chemoradiotherapeutic response, with AUC values of 0.52, 0.33, and 0.32, respectively (all P > 0.05).

Typical PET images from CR and non-CR groups are shown in Figs. 3 and 4, respectively.